Forum Moderators: open
For example, take "<some-place> web design". My company used to be number 1 in my own damn home town before the Google changes in Sept. Now, some results simply have nothing to do with "web design" as a service. All someone has to do now, for example a <widget> web site, is to put the words..."Come take a look at our <widgets> on our WEB SITE and see our wonderful <widget> DESIGNS, stop by our shop in <some-place>" and boom, they get listed in Google under web design search and they don't even do web site design.
The search results simply are now just looking at any random word and no longer looking at the actual content of the pages.
It took me two years to finally get good positioning in the search engines and Google had to go and mess it up.
I think MSN is a much more accurate search engine. At least there they show you real content with accurate results.
[edited by: ciml at 2:12 pm (utc) on Jan. 17, 2003]
[edit reason] Widget-ised. [/edit]
Well i guess that google is not putting enough weight on words that appear together. But im sure everybody has thought of that.
I know Alltheweb for example has a pretty nifty way of dealing with this, though it has some drawbacks of its own that involves "rewriting" common phrases, leeting you know in the sERPs what it has done, and allowing you one clik access to change it back again.
BTW, to see how much better ideal use of quotations matters:
[google.com...]
I am sure most people do not use quotes in their searchs and i am saying this based on a custom script on my site that keep track of the exact search term, with or without quotes, etc.
The reality is that Google results are often much better when quotes are used. So it appears to me that Google is expecting people to put the quotes in the search phrase.
Educating "Joe Public" is not an easy thing to do. It will be a long time before people realize how to operate a search engine in the way the search engine desires. They are more likely to switch to a different search engine in the meantime.
I believe Google needs to be a little smarter and take account of how people actually use it. The simple solution is to increase the weighting of the proximity of the words in search terms.
Try a test. Think of 10+ random, but reasonably popular multiple word search phases. Try "normal" unquoted searches for these phases, and then try the quoted versions. If the quoted versions produce more relevant results why isn't the search engine smart enough to assume that proximity/backlink anchor text is more important than PR?
Google's algo is assuming the intelligence should sit with the user (i.e. they should use quotes), a good search engine should compensate for this fact and be able to deliver good results based upon the entire phrase the user entered.
I did the above test, my results produced a Google accuracy of 97% using quotes and a 76% accuracy using the non quoted phrases.
If these stats seem close to others trying a random sample, then can anyone explain why Google wouldn't want to achieve the additional 21% accuracy level? Selling Adwords was the only explanation/strategy I could come up with!
They think the average engine is as smart as the computer on StarTrek or even AskJeeves when it is parsing results! We may know it "ain't so", but the average user doesn't.
As for the problem of where a piece of software should add "quotes", an average math or stats major could come up with the answer, the PHD's at Google should have no trouble with finding a good solution.
Monkscuba, I wish the average person looking for a web design company was as savvy as you think. But, in reality they aren't. Most web sites are not made for large corporate entities with SE friendly IT professionals, they are made for the "Mom and Pop" type businesses that know they need to be on the web. The "Joe Public" types. These folks typically don't know how to use a search engine to its max.
Search quality is many things to many people. It depends greatly on geography as well as other "influencing" factors.
At the end of the day Google became popular because it produced good, fast, clear and simple relevant results. It will only remain popular on that basis. Competitors are out there waiting for it to make mistakes, and I personally believe it is in the process of doing so!
So yes, its important to come up in Google as a web design service in your area for new companies trawling the mom-and-pop market and i guess some of these SERPs are showing that it is difficult to do.
Of course the above are generalizations, but still useful ones. Im sure many web design companies have sold a lot of new accounts from a web lead.
1. Google no longer favors "exact phrase" search matches; in fact, may penalise pages if title = "exact phrase" and "exact phrase" is found in meta tags, heading and emphasized body copy.
2. Google somehow places much higher emphasis on the geographic reference of the search phrase (the term "New York" has much more weight when calculating results than the term "Green Widgets").
So, to my way of thinking, the results returned for Marcia's search phrases are perfectly "Google-ish", reflecting the algo changes since summer. The fact that they are off-target says to me that it's hard to de-emphasize "exact phrase" matches, emphasize only part of a multiple word search phrase and still have a great index.
As RFG points out, including phrases (quotation marks) in the search often leads to much higher relevance. As Marcia points out, not many people use quotes. Google's solution was to give a proximity bonus by default.
If AltaVista users had tended to type {web NEAR site NEAR design NEAR new NEAR york NEAR city} then there wouldn't have been quite such a need for Google. In my opinion the default proximity bonus was as important for multiword phrases as PageRank was for single word phrases.
Quite a few of us have seen what we believe to be a reduction in the proximity bonus. Maybe Google needed to do that to fight spam, but their ability to give us the most relevant pages near the top of the listings did suffer IMO.